[Predictive value of vascular health indicators on newly cardiovascular events: Preliminary validation of Beijing vascular health stratification system].

OBJECTIVE

To explore the predictive value of carotid femoral artery pulse wave velocity (CF-PWV), carotid radial artery pulse wave velocity (CR-PWV), cardio-ankle vascular index (CAVI), and ankle brachial index (ABI) on coronary heart disease (CHD) and cerebral infarction (CI), and the preliminary validation of Beijing vascular health stratification (BVHS).

METHODS

Subjects with at least 2 in-patient records were included into the study between 2010 and 2017 from Vascular Medicine Center of Peking University Shougang Hospital. Subjects with CHD or CI, and without data of vascular function at baseline were excluded. Eventually, 467 subjects free of CHD [cohort 1, mean age: (63.4±12.3) years, female 42.2%] and 658 subjects free of CI [cohort 2, mean age: (64.3±12.2) years, female 48.7%] at baseline were included. The first in-patient records were as the baseline data, the second in-patient records were as a following-up data. Cox proportional hazard regression was used to establish the predictive models of CHD or CI derived from BVHS by multivariable-adjusted analysis.

RESULTS

The median follow-up time of cohort 1 and cohort 2 was 1.9 years and 2.1 years, respectively. During the follow-up, 164 first CHD events occurred in cohort 1 and 117 first CI events occurred in cohort 2. Four indicators were assessed as continuous variables simultaneously by multivariable-adjusted analysis. In cohort 1, CF-PWV, CR-PWV, ABI, and CAVI reached statistical significance in the multivariable-adjusted models (<0.05). In cohort 2, only CAVI (<0.05) was of statistical significance. In addition, the higher CF-PWV became a protector of CHD or CI (<0.05). The prediction value of BVHS reached the statistical significance for CHD and CI in the unadjusted models (all <0.05), however, BVHS could only predict the incidence of CHD (<0.05), but not the incidence of CI (>0.05) in the multivariable-adjusted models. CF-PWV, CR-PWV, ABI, and CAVI were associated factors of CHD independent of each other (<0.05), only CAVI (<0.05) was the risk factor of CI independent of the other three.

CONCLUSION

The different vascular indicators might have different effect on CHD or CI. CAVI might be a stable predictor of both CHD and CI. Higher baseline CF-PWV was not necessarily a risk factor of CHD or CI because of proper vascular health management. BVHS was a potential factor for the prediction of CHD, and further research is needed to explore the prediction value for CI.